CN109948398A - The image processing method and panorama parking apparatus that panorama is parked - Google Patents
The image processing method and panorama parking apparatus that panorama is parked Download PDFInfo
- Publication number
- CN109948398A CN109948398A CN201711387606.5A CN201711387606A CN109948398A CN 109948398 A CN109948398 A CN 109948398A CN 201711387606 A CN201711387606 A CN 201711387606A CN 109948398 A CN109948398 A CN 109948398A
- Authority
- CN
- China
- Prior art keywords
- birds
- eye
- overlapping region
- panorama
- camera
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000003672 processing method Methods 0.000 title claims abstract description 22
- 230000000007 visual effect Effects 0.000 claims abstract description 9
- 238000012937 correction Methods 0.000 claims description 42
- 235000004522 Pentaglottis sempervirens Nutrition 0.000 claims description 14
- 239000011159 matrix material Substances 0.000 claims description 11
- 238000010586 diagram Methods 0.000 description 9
- 238000000034 method Methods 0.000 description 9
- 230000006870 function Effects 0.000 description 8
- 238000003860 storage Methods 0.000 description 5
- 241000251468 Actinopterygii Species 0.000 description 4
- 210000003128 head Anatomy 0.000 description 4
- 230000008569 process Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000005286 illumination Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 241000736199 Paeonia Species 0.000 description 1
- 235000006484 Paeonia officinalis Nutrition 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 210000003733 optic disk Anatomy 0.000 description 1
- 238000003825 pressing Methods 0.000 description 1
- 230000001902 propagating effect Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Abstract
The invention discloses a kind of image processing methods that panorama is parked comprising: the multiple images in all angulars field of view of vehicle periphery are obtained using multiple cameras, two neighboring camera acquired image has certain overlapping region in multiple cameras;Multiple images are handled to obtain multiple birds-eye views respectively;The characteristic point of the overlapping region of multiple birds-eye views is extracted respectively and characteristic matching is carried out to characteristic point, and multiple birds-eye views are projected into the same view visual angle;The optimal stitching line of the overlapping region in multiple birds-eye views is established, and is spliced multiple birds-eye views according to optimal stitching line to obtain the first panorama sketch;With graph cut is carried out to the first panorama sketch to eliminate splicing trace to generate the second panorama sketch.In addition, the invention also discloses a kind of panorama parking apparatus.The image processing method and panorama parking apparatus that panorama of the invention is parked avoid ghost phenomenon and eliminate splicing trace when generating panorama birds-eye view, guarantee panorama birds-eye view brightness uniformity.
Description
Technical field
The present invention relates to technical field of automotive electronics, in particular to a kind of image processing method that panorama is parked and panorama pool
Vehicle device.
Background technique
In the related art, by being mounted on vehicle body, all around four road fish-eye cameras acquire panoramic parking system in real time
Multiple images are synthesized a complete birds-eye view panoramic picture by image procossing by the image around vehicle body.Pass through panorama sketch
Picture, driver can understand the position of vehicle's surroundings barrier and pedestrian and its at a distance from vehicles, avoid vision dead zone and
Dead angle, panoramic parking system can effectively avoid vehicle wiping from touching accident, improve the safe maneuverability of vehicle.
However, some panoramic parking systems only simply by the image mosaic of all directions together, this causes image to exist
There is apparent splicing gap within the scope of overlapping region;In addition, some panoramic parking systems directly merge overlapping region, but
It fogs since object parts overlapping will lead to final panoramic picture, that is, there is ghost phenomenon, so as to cause poor user experience.
Summary of the invention
Embodiments of the present invention provide a kind of image processing method that panorama is parked and panorama parking apparatus.
The image processing method that the panorama of embodiment of the present invention is parked, comprising the following steps:
The multiple images in all angulars field of view of vehicle periphery, phase in the multiple camera are obtained using multiple cameras
Adjacent two camera acquired images have certain overlapping region;
Described multiple images are handled to obtain multiple birds-eye views respectively;
The characteristic point of the overlapping region of the multiple birds-eye view is extracted respectively and characteristic matching is carried out to the characteristic point, and
The multiple birds-eye view is projected into the same view visual angle;
The optimal stitching line of the overlapping region in the multiple birds-eye view is established, and will according to the optimal stitching line
The multiple birds-eye view is spliced to obtain the first panorama sketch;With
Graph cut is carried out to first panorama sketch to eliminate splicing trace to generate the second panorama sketch.
The image processing method that the panorama of embodiment of the present invention is parked, by establishing optimal stitching line for multiple birds-eye views
Spliced, avoid directly merge and generate ghost phenomenon, and using graph cut to optimal stitching line merged with
Eliminate the splicing trace of overlapping region.In this way, guaranteeing panorama birds-eye view brightness uniformity, to improve user experience.
In some embodiments, the camera is fish-eye camera, and the processing described multiple images to obtain respectively
Obtaining multiple birds-eye views includes:
The multiple fish-eye camera is demarcated, the correction parameter of each fish-eye camera is obtained and utilizes institute
Correction parameter processing described multiple images are stated to obtain multiple correction charts respectively;
According to homography matrix, the multiple correction chart is handled using bilinear interpolation described to obtain respectively
Multiple birds-eye views.
In some embodiments, the optimal stitching line packet for establishing the overlapping region in the multiple birds-eye view
It includes:
Using each pixel of the frame of the overlapping region as the starting point of suture;
Multiple sutures are determined from the starting point of the suture, and calculate the overall strength of the pixel of each suture
Value;With
Select the smallest suture of total intensity value as the optimal stitching line from the multiple suture.
In some embodiments, the starting point from the suture determines that multiple sutures include: from the starting point
The smallest pixel of intensity value is as the suture in adjacent with the starting point and next pixel column positioned at the starting point of lookup
The current point of line searches the smallest pixel of intensity value in next pixel column that is adjacent with the current point and being located at the current point
The next point as the suture is put until exceeding the overlapping region.
In some embodiments, the intensity value calculation formula of each pixel are as follows:
E (x, y)=Ecolor(x,y)2+Egeometry(x,y);
Wherein, E (x, y) indicates same in the overlapping region of two birds-eye views obtained by the two neighboring camera
The intensity value of pixel at one coordinate, Ecolor(x, y) is got a bird's eye view described in indicating two obtained as the two neighboring camera
The difference of the color value of pixel in the overlapping region of figure at same coordinate, Egeometry(x, y) indicates described to be taken the photograph by two neighboring
The structure difference of pixel in the overlapping region of two birds-eye views obtained as head at same coordinate.
In some embodiments, definition is located at the overlapping region part difference of the optimal stitching line two sides
For the first sub- overlapping region and the second sub- overlapping region, described image and second son where the first sub- overlapping region
Described image where overlapping region belongs to two birds-eye views that two neighboring camera obtains, described according to the best suture
The multiple birds-eye view is spliced to obtain the first panorama sketch to include: to fill institute respectively in the two sides of the optimal stitching line by line
The first sub- overlapping region and the second sub- overlapping region are stated to splice to obtain first panorama sketch.
The panorama parking apparatus of embodiment of the present invention, comprising:
Multiple cameras, the multiple camera are used to obtain the multiple images in all angulars field of view of vehicle periphery, institute
Stating two neighboring camera acquired image in multiple cameras has certain overlapping region;
Image processing unit, described image processing unit is for handling described multiple images to obtain multiple get a bird's eye view respectively
Figure;
Image mosaic unit, described image concatenation unit are used to extract the spy of the overlapping region of the multiple birds-eye view respectively
Sign point simultaneously carries out characteristic matching to the characteristic point, and the multiple birds-eye view is projected to the same view visual angle, and establish
The optimal stitching line of the overlapping region in the multiple birds-eye view, and got a bird's eye view according to the optimal stitching line by the multiple
Figure is spliced to obtain the first panorama sketch;With
Seam integrated unit, the seam integrated unit are used to carry out graph cut to first panorama sketch to spell to eliminate
Lapping defect mark is to generate the second panorama sketch.
The panorama parking apparatus of embodiment of the present invention is spliced multiple birds-eye views by establishing optimal stitching line,
The ghost phenomenon for directly merging and generating is avoided, and optimal stitching line is merged to eliminate overlay region using graph cut
The splicing trace in domain.In this way, guaranteeing panorama birds-eye view brightness uniformity, to improve user experience.
In some embodiments, the camera is fish-eye camera, and described image processing unit includes:
Distortion correction unit, the distortion correction unit obtain every for demarcating to the multiple fish-eye camera
The correction parameter of a fish-eye camera simultaneously handles described multiple images using the correction parameter to obtain multiple schools respectively
Positive figure;With
Birds-eye view processing unit, the birds-eye view processing unit are used to utilize bilinear interpolation according to homography matrix
The multiple correction chart is handled to obtain the multiple birds-eye view respectively.
In some embodiments, described image concatenation unit is used for:
Using each pixel of the frame of the overlapping region as the starting point of suture;
Multiple sutures are determined from the starting point of the suture, and calculate the overall strength of the pixel of each suture
Value;With
Select the smallest suture of total intensity value as the optimal stitching line from the multiple suture.
In some embodiments, the intensity value calculation formula of each pixel are as follows:
E (x, y)=Ecolor(x,y)2+Egeometry(x,y);
Wherein, E (x, y) indicates same in the overlapping region of two birds-eye views obtained by the two neighboring camera
The intensity value of pixel at one coordinate, Ecolor(x, y) is got a bird's eye view described in indicating two obtained as the two neighboring camera
The difference of the color value of pixel in the overlapping region of figure at same coordinate, Egeometry(x, y) indicates described to be taken the photograph by two neighboring
The structure difference of pixel in the overlapping region of two birds-eye views obtained as head at same coordinate.
Additional aspect and advantage of the invention will be set forth in part in the description, and will partially become from the following description
Obviously, or practice through the invention is recognized.
Detailed description of the invention
Above-mentioned and/or additional aspect and advantage of the invention is from combining in description of the following accompanying drawings to embodiment by change
It obtains obviously and is readily appreciated that, in which:
Fig. 1 is the flow diagram for the image processing method that the panorama of embodiment of the present invention is parked;
Fig. 2 is another flow diagram for the image processing method that the panorama of embodiment of the present invention is parked;
Fig. 3 is the module diagram of the panorama parking apparatus of embodiment of the present invention;
Fig. 4 is another module diagram of the panorama parking apparatus of embodiment of the present invention;
Fig. 5 is the schematic diagram of the scaling board of embodiment of the present invention;
Fig. 6 is the placement schematic diagram of the scaling board of embodiment of the present invention;
Fig. 7 is the another flow diagram for the image processing method that the panorama of embodiment of the present invention is parked;
Fig. 8 is the partial routine schematic diagram for establishing optimal stitching line of embodiment of the present invention;
Fig. 9 is the birds-eye view splicing schematic diagram of embodiment of the present invention.
Main element symbol description:
Panorama parking apparatus 100, camera 110, image processing unit 120, distortion correction unit 122, birds-eye view processing
Unit 124, image mosaic unit 130, seam integrated unit 140, overlapping region 10, first sub- the 12, second son of overlapping region weight
Folded region 14, optimal stitching line 20.
Specific embodiment
Embodiments of the present invention are described below in detail, the embodiment of the embodiment is shown in the accompanying drawings, wherein
Same or similar label indicates same or similar element or element with the same or similar functions from beginning to end.Lead to below
It crosses the embodiment being described with reference to the drawings to be exemplary, for explaining only the invention, and should not be understood as to limit of the invention
System.
Please refer to Fig. 1-Fig. 4, the image processing method that the panorama of embodiment of the present invention is parked, comprising:
Step S110: obtaining the multiple images in all angulars field of view of vehicle periphery using multiple cameras 110, multiple to take the photograph
As two neighboring 110 acquired image of camera has certain overlapping region in first 110.
In embodiments of the present invention, camera 110 is fish-eye camera.In one embodiment, fish-eye camera
Quantity be four, respectively preceding fish-eye camera, rear fish-eye camera, left fish-eye camera and right fish-eye camera.
It is appreciated that fish-eye camera angular field of view is big, it can be with a wide range of scenery of shooting at close range.Four fish-eye cameras
The image and two neighboring fish-eye camera acquired image obtained in all angulars field of view of vehicle periphery enough has centainly
Overlapping region.Certainly, in another embodiment, other kinds of camera can also be used in camera 110, need only guarantee to take the photograph
The image in all angulars field of view of vehicle periphery and 110 acquired image of two neighboring camera can be obtained have as first 110
There is certain overlapping region.The quantity of camera 110 is it is not limited here.The quantity of camera 110 can also be according to actual needs
Specific setting.
Specifically, preceding fish-eye camera is mounted on the position of Chinese herbaceous peony, for acquiring the image of vehicle front region (before calling in the following text
View);Fish-eye camera is mounted on tailstock position afterwards, for acquiring the image (calling rearview in the following text) of rear view of vehicle region;Left fish
Eye imaging head is mounted on left-hand mirror, for acquiring the image (calling left view in the following text) of vehicle left area;Right fish-eye camera peace
Mounted in right rear view mirror, for acquiring the image (calling right view in the following text) in vehicle right region.When installing fish-eye camera, in flake
The face position of camera can place scaling board, guarantee that fish-eye camera can clearly take all angle points of scaling board;?
Two adjacent fish-eye cameras are formed by the overlapping region in image and place scaling board, guarantee adjacent two flakes camera shooting
Head can clearly take all angle points of scaling board.Referring to Fig. 5, scaling board can be chessboard calibration plate.Each grid
Four vertex are the angle point of scaling board.
Referring to Fig. 6, being put in the face position of each fish-eye camera and the shooting overlapping region of adjacent fish-eye camera
Scaling board is set, scaling board has eight altogether.Fish-eye camera acquisition includes the multiple images of scaling board.Multiple images include forward sight
Figure, rearview, left view and right view.
Step S120: processing multiple images to obtain multiple birds-eye views respectively.
When camera 110 is fish-eye camera, step S120: processing multiple images to obtain multiple birds-eye view packets respectively
It includes:
Step S122: demarcating multiple fish-eye cameras, obtains the correction parameter of each fish-eye camera and utilization
Correction parameter handles multiple images to obtain multiple correction charts respectively;
Step S124: according to homography matrix, multiple correction charts are handled to obtain respectively using bilinear interpolation
Obtain multiple birds-eye views.
It is appreciated that image (calling fish eye images in the following text) captured by fish-eye camera can generate distortion, therefore, it is necessary to figure
As carrying out distortion correction.Multiple fish-eye cameras are mounted on the different location of vehicle, and each fish-eye camera is relative to horizontally
The angle in face is different.Therefore, in order to obtain the panorama sketch on ground, need for multiple images to be transformed to multiple birds-eye views, then carry out
Splicing.
Specifically, according to the angular coordinate and distortion model of the scaling board for being located at face position in fish eye images, certain
Fish-eye camera parameter is traversed in range, obtains the mark for being located at face position under different fish-eye camera Parameter Conditions in correction chart
The coordinate of the angle point of fixed board.The angular coordinate and ideal that the scaling board of face position will be located in the correction chart obtained every time are demarcated
The angular coordinate of plate compares, and the correction parameter of fish-eye camera is obtained when angular coordinate difference minimum.In this way, to four
Fish-eye camera is successively demarcated, and is obtained the correction parameter of each fish-eye camera and is handled multiple flakes using correction parameter
Image is to obtain multiple correction charts.In the present embodiment, distortion model be equidistant projection model, camera parameter include focal length,
Optical center position and unit pixel the ratio of width to height etc..Here, being not limited to the above embodiments, other suitable distortion correction sides can be used
Formula.
Then, it according to the angular coordinate and correction parameter of the scaling board for being located at face position in fish eye images, is corrected
It is located at the angular coordinate of the scaling board of face position in figure.Assuming that being located at the angular coordinate of the scaling board of face position in correction chart
For (X1, Y1), the corresponding angular coordinate of ideal scaling board is (X2, Y2) in top view, and relationship between the two meets:
Wherein, (X1, Y1) indicates all angular coordinates for being located at the scaling board of face position in correction chart, (X2, Y2) table
Show the corresponding all angular coordinates of ideal scaling board in top view.The angular coordinate of ideal scaling board is ideal calibration in top view
Angular coordinate of the plate on level ground.Using in known correction chart be located at face position scaling board angular coordinate and bow
The corresponding angular coordinate of ideal scaling board, calculates homography matrix in view.To front view, rearview, left view and right view
Figure calculates separately corresponding homography matrix.According to corresponding homography matrix, bird before being generated respectively using bilinear interpolation
It looks down from a height figure, rear birds-eye view, left birds-eye view and right birds-eye view.Preceding birds-eye view, rear birds-eye view, left birds-eye view and right birds-eye view constitute multiple
Birds-eye view.
Step S130: extracting the characteristic point of the overlapping region of multiple birds-eye views respectively and carries out characteristic matching to characteristic point,
And multiple birds-eye views are projected into the same view visual angle.
It is appreciated that in step s 130, preceding birds-eye view and left birds-eye view have an overlapping region, preceding birds-eye view and right bird
Figure of looking down from a height has an overlapping region, and rear birds-eye view and left birds-eye view have an overlapping region, and rear birds-eye view and right birds-eye view have one
Overlapping region.In the present embodiment, the characteristic point of overlapping region is the angle point for being located at the scaling board of overlapping region.
In one embodiment, multiple birds-eye views can be projected to forward sight angle of field.Before being projected to below with left birds-eye view
For view visual angle, step S130 is illustrated.Assuming that (X3, Y3) is the scaling board for being located at overlapping region in preceding birds-eye view
Angular coordinate, (X4, Y4) are the corresponding angular coordinates of scaling board for being located at overlapping region in left birds-eye view.Left birds-eye view is carried out
Projective transformation meets following formula to angle of field is before got a bird's eye view:
Wherein, it is located at all angular coordinates of the scaling board of overlapping region, (X4, Y4) before (X3, Y3) is indicated in birds-eye view
Indicate the corresponding all angular coordinates of scaling board for being located at overlapping region in left birds-eye view.According to preceding birds-eye view and left birds-eye view pair
The scaling board angular coordinate answered calculates homography matrix.According to homography matrix, left birds-eye view is thrown using bilinear interpolation
Shadow is to before getting a bird's eye view angle of field.As example, image registration is carried out to two birds-eye views that two neighboring fish-eye camera obtains, it will
All birds-eye views get a bird's eye view angle of field before projecting to.Particularly, it since rear birds-eye view and preceding birds-eye view do not have overlapping region, needs
Rear birds-eye view is first projected into left birds-eye view and angle of field is got a bird's eye view on the right side, reprojection is to before getting a bird's eye view angle of field.
Step S140: the optimal stitching line of the overlapping region in multiple birds-eye views is established, and will be more according to optimal stitching line
A birds-eye view is spliced to obtain the first panorama sketch.
It is appreciated that there are four overlapping regions in four birds-eye views in step S140.Therefore, it is necessary to establish four most
Four birds-eye views could be spliced into panorama birds-eye view by good suture.
In some embodiments, referring to Fig. 7, establishing the optimal stitching line packet of the overlapping region in multiple birds-eye views
It includes:
Step S142: using each pixel of the frame of overlapping region as the starting point of suture;
Step S144: determining multiple sutures from the starting point of suture, and calculate each suture pixel it is total strong
Angle value;With
Step S146: select the smallest suture of total intensity value as optimal stitching line from multiple sutures.
In this way, by comparing the total intensity value of multiple sutures, to select optimal stitching line.
In some embodiments, determine that multiple sutures include: to search and starting point phase from starting point from the starting point of suture
Current point of the smallest pixel of intensity value as suture in next pixel column that is adjacent and being located at starting point, is searched and current point phase
The smallest pixel of intensity value is as the next point of suture until beyond overlapping in next pixel column that is adjacent and being located at current point
Region.
In some embodiments, the intensity value calculation formula of each pixel are as follows:
E (x, y)=Ecolor(x,y)2+Egeometry(x,y);
Wherein, E (x, y) indicates same seat in the overlapping region of two birds-eye views obtained by two neighboring camera 110
The intensity value of pixel at mark, Ecolor(x, y) indicates the overlay region of two birds-eye views obtained by two neighboring camera 110
The difference of the color value of pixel in domain at same coordinate, Egeometry(x, y) indicates obtained by two neighboring camera 110 two
The structure difference of pixel in the overlapping region of a birds-eye view at same coordinate.
It is appreciated that two birds-eye views (calling two neighboring birds-eye view in the following text) that two adjacent fish-eye cameras obtain are deposited
In overlapping region, therefore, the registration accuracy of the pixel on optimal stitching line is highest in overlapping region.The best suture of composition
The pixel of line should meet: in color, the difference of color value of the pixel of suture on two neighboring birds-eye view is minimum;?
In structure, structure of the pixel on two neighboring birds-eye view on suture is most like.In this way, pressing formula E (x, y)=Ecolor
(x,y)2+Egeometry(x, y) establishes optimal stitching line and meets two above condition.Structure difference can carry out ladder by gradient operator
Degree, which calculates, to be obtained.Wherein, the gradient operator in the direction x and the direction y is respectively
Specifically, in one example, the establishment process of optimal stitching line is as follows: referring to Fig. 8, assuming overlapping region packet
5*5 pixel is included, circle indicates pixel in figure.Pixel 11 indicates the pixel for being located at the 1st pixel column of the 1st pixel column,
And so on.
The frame of overlapping region is the frame that the 1st pixel column, the 1st pixel column, the 5th pixel column and the 5th pixel column are constituted.?
In this example, using each pixel in the 1st pixel column of overlapping region and the 1st pixel column as the starting point of suture.
Multiple sutures are determined from the starting point of suture, in fig. 8, left figure expression has determined the starting point of 9 sutures, point
It Wei not pixel 11,12,13,14,15,21,31,41 and 51;Right figure indicates that suture determines the extension after starting point.In right figure
In, for example, three pixels 21 adjacent with the starting point from next pixel column that starting point 12 issues three line segments direction starting points,
22 and 23, current point of the smallest pixel 23 of intensity value as suture is found, real segment indicates that directional intensity value is minimum
Pixel;Three line segments are issued from current point 23 to be directed toward adjacent with the current point 23 three in next pixel column of current point 23
A pixel 32,33 and 34 finds next point of the smallest pixel 33 of intensity value as suture;In this way, continuing down
One pixel column searches the smallest pixel of intensity value until the 5th pixel column.In right figure, illustrate only with pixel 12 as starting point
Suture and be the suture of starting point with pixel 21.It is every to expand to next pixel column during determining multiple sutures
A pixel is opened up, the total intensity value of a suture is calculated.The intensity value of pixel is according to formula E (x, y)=Ecolor(x,y
)2+Egeometry(x, y) is calculated.Total intensity value is the summation of the intensity value of all pixels point on suture.
Select the smallest suture of total intensity value as optimal stitching line from 9 sutures.For example, in the right figure of Fig. 8
In, it may compare the total intensity value of suture 21-31-42-53 and the total intensity value of suture 12-23-33-44-54, determine total strong
The smallest suture of angle value is as optimal stitching line, and e.g. suture 12-23-33-44-54 is as optimal stitching line.
Multiple birds-eye views are spliced to obtain the first panorama sketch according to optimal stitching line specific as follows: referring to Fig. 9, with
Preceding birds-eye view and for getting a bird's eye view the splicing of left birds-eye view of angle of field before projecting to, before solid box indicates in the figure birds-eye view and
The overlapping region 10 of the left birds-eye view of angle of field is got a bird's eye view before projecting to.Definition is located at the overlay region of 20 two sides of optimal stitching line
10 part of domain is respectively the first sub- overlapping region 12 and the second sub- overlapping region 14.Image where first sub- overlapping region 12 is
Preceding birds-eye view, the image where the second sub- overlapping region 14 is the left birds-eye view that angle of field is got a bird's eye view before projecting to.In overlapping region
Sub- 12 fills of overlapping region of the first of the preceding birds-eye view in the right side of optimal stitching line 20, optimal stitching line in overlapping region 10
It fills angle of field is got a bird's eye view in 20 left side the second sub- overlapping region 14 before being projected to left birds-eye view.In this way, in the same way
Four birds-eye views at same view visual angle are spliced into the first panorama sketch.
Step S150: graph cut is carried out to the first panorama sketch to eliminate splicing trace to generate the second panorama sketch.
It is appreciated that in step S140, the first panorama sketch obtained according to optimal stitching line direct splicing splices trace
Obviously.Further, since fish-eye camera shooting direction and angle are different, two neighboring fish-eye camera illumination there may be
Difference.Therefore, graph cut is carried out to the overlapping region in the first panorama sketch, to eliminate splicing trace, uniform illumination, thus raw
At the second panorama sketch of brightness uniformity.
The image processing method that the panorama of embodiment of the present invention is parked, by establishing optimal stitching line for multiple birds-eye views
Spliced, avoid directly merge and generate ghost phenomenon, and using graph cut to optimal stitching line merged with
Eliminate the splicing trace of overlapping region.In this way, guaranteeing panorama birds-eye view brightness uniformity, to improve user experience.
It is appreciated that the image processing method that the panorama of embodiment of the present invention is parked can be used for assisting the driving of vehicle.
Referring to Fig. 3, the panorama parking apparatus 100 of embodiment of the present invention, including multiple cameras 110, image procossing
Unit 120, image mosaic unit 130 and seam integrated unit 140.Multiple cameras 110 are for obtaining all views of vehicle periphery
Multiple images in angular region, two neighboring 110 acquired image of camera has certain overlapping in multiple cameras 110
Region.Image processing unit 120 is for handling multiple images to obtain multiple birds-eye views respectively.Image mosaic unit 130 is used for
The characteristic point of the overlapping region of multiple birds-eye views is extracted respectively and characteristic matching is carried out to characteristic point, and multiple birds-eye views are projected
To the same view visual angle and the optimal stitching line for establishing the overlapping region in multiple birds-eye views, and will be more according to optimal stitching line
A birds-eye view is spliced to obtain the first panorama sketch.Seam integrated unit 140 be used for the first panorama sketch carry out graph cut with
Splicing trace is eliminated to generate the second panorama sketch.
In other words, the image processing method that the panorama of embodiment of the present invention is parked can be by embodiment of the present invention
Panorama parking apparatus 100 is realized.Wherein step S110 can realize that step S120 can be by image by multiple cameras 110
It manages unit 120 and realizes that step S130 and step S140 can be realized that step S150 can be by seam by image mosaic unit 130
Integrated unit 140 is realized.
The panorama parking apparatus 100 of embodiment of the present invention is spelled multiple birds-eye views by establishing optimal stitching line
It connects, avoids the ghost phenomenon for directly merging and generating, and merged to optimal stitching line to eliminate weight using graph cut
The splicing trace in folded region.In this way, guaranteeing panorama birds-eye view brightness uniformity, to improve user experience.
It is appreciated that the explanation and beneficial effect of the embodiment of the above-mentioned image processing method parked to panorama
Panorama parking apparatus 100 suitable for embodiment of the present invention is no longer developed in details herein to avoid redundancy.
In some embodiments, referring to Fig. 4, camera 110 is fish-eye camera, image processing unit 120 includes
Distortion correction unit 122 and birds-eye view processing unit 124.Distortion correction unit 122 is for marking multiple fish-eye cameras
It is fixed, obtain the correction parameter of each fish-eye camera and using correction parameter processing multiple images to obtain multiple corrections respectively
Figure.Birds-eye view processing unit 124 is used for according to homography matrix, using bilinear interpolation to multiple correction charts handled with
Multiple birds-eye views are obtained respectively.
In other words, step S122 can be realized by distortion correction unit 122, and step S124 can be handled single by birds-eye view
Member 124 is realized.
In some embodiments, image mosaic unit 130 is used for:
Using each pixel of the frame of overlapping region as the starting point of suture;
Multiple sutures are determined from the starting point of suture, and calculate the total intensity value of the pixel of each suture;With
Select the smallest suture of total intensity value as optimal stitching line from multiple sutures.
In other words, step S142, step S144 and step S146 are realized by image mosaic unit 130.
In some embodiments, image mosaic unit 130 is used to search adjacent with starting point from starting point and is located at starting point
Current point of the smallest pixel of intensity value as suture in next pixel column searches adjacent with current point and is located at current point
Next pixel column in the smallest pixel of intensity value as suture next point until exceed overlapping region.
In this way, establishing multiple sutures so that the smallest suture of subsequently selected total intensity value is as optimal stitching line.
In some embodiments, it is respectively the first son that definition, which is located at the overlapping region position of optimal stitching line two sides,
Overlapping region and the second sub- overlapping region.The image category where image and the second sub- overlapping region where first sub- overlapping region
In two birds-eye views that two neighboring camera 110 obtains.Image mosaic unit 130 is used for: in the two sides of optimal stitching line point
Not Tian Chong the first sub- overlapping region and the second sub- overlapping region to splice to obtain the first panorama sketch.
It is appreciated that the two sides of optimal stitching line are used respectively two acquired in two neighboring camera 110 in overlapping region
The content of a birds-eye view is filled, to avoid overlapped object, i.e. ghost.
In the description of embodiments of the present invention, term " first ", " second " are used for description purposes only, and cannot understand
For indication or suggestion relative importance or implicitly indicate the quantity of indicated technical characteristic.Define as a result, " first ",
The feature of " second " can explicitly or implicitly include one or more feature.In embodiments of the present invention
In description, the meaning of " plurality " is two or more, unless otherwise specifically defined.
In the description of embodiments of the present invention, it should be noted that unless otherwise clearly defined and limited, term
" installation ", " connected ", " connection " shall be understood in a broad sense, for example, it may be fixedly connected, may be a detachable connection or one
Connect to body;It can be mechanical connection, be also possible to be electrically connected or can mutually communicate;It can be directly connected, can also lead to
It crosses intermediary to be indirectly connected, can be the connection inside two elements or the interaction relationship of two elements.For ability
For the those of ordinary skill in domain, can understand as the case may be above-mentioned term in embodiments of the present invention specifically contain
Justice.
In the description of this specification, reference term " embodiment ", " some embodiments ", " schematically implementation
The description of mode ", " example ", specific examples or " some examples " etc. means the tool described in conjunction with the embodiment or example
Body characteristics, structure, material or feature are contained at least one embodiment or example of the invention.In the present specification,
Schematic expression of the above terms are not necessarily referring to identical embodiment or example.Moreover, the specific features of description, knot
Structure, material or feature can be combined in any suitable manner in any one or more embodiments or example.
Any process described otherwise above or method description are construed as in flow chart or herein, and expression includes
It is one or more for realizing specific logical function or process the step of executable instruction code module, segment or portion
Point, and the range of the preferred embodiment of the present invention includes other realization, wherein can not press shown or discussed suitable
Sequence, including according to related function by it is basic simultaneously in the way of or in the opposite order, Lai Zhihang function, this should be of the invention
Embodiment person of ordinary skill in the field understood.
Expression or logic and/or step described otherwise above herein in flow charts, for example, being considered use
In the order list for the executable instruction for realizing logic function, may be embodied in any computer-readable medium, for
Instruction execution system, device or equipment (such as computer based system, including the system of processing module or other can be from instruction
Execute system, device or equipment instruction fetch and the system that executes instruction) use, or combine these instruction execution systems, device or
Equipment and use.For the purpose of this specification, " computer-readable medium " can be it is any may include, store, communicating, propagating or
Transfer program uses for instruction execution system, device or equipment or in conjunction with these instruction execution systems, device or equipment
Device.The more specific example (non-exhaustive list) of computer-readable medium include the following: there are one or more wirings
Electrical connection section (electronic device), portable computer diskette box (magnetic device), random access memory (RAM), read-only memory
(ROM), erasable edit read-only storage (EPROM or flash memory), fiber device and portable optic disk is read-only deposits
Reservoir (CDROM).In addition, computer-readable medium can even is that the paper that can print described program on it or other are suitable
Medium, because can then be edited, be interpreted or when necessary with it for example by carrying out optical scanner to paper or other media
His suitable method is handled electronically to obtain described program, is then stored in computer storage.
It should be appreciated that each section of embodiments of the present invention can be with hardware, software, firmware or their combination come real
It is existing.In the above-described embodiment, multiple steps or method can be with storages in memory and by suitable instruction execution system
The software or firmware of execution is realized.For example, if realized with hardware, in another embodiment, ability can be used
Any one of following technology or their combination well known to domain is realized: being had for realizing logic function to data-signal
The discrete logic of logic gates, the specific integrated circuit with suitable combinational logic gate circuit, programmable gate array
(PGA), field programmable gate array (FPGA) etc..
Those skilled in the art are understood that realize all or part of step that above-described embodiment method carries
It suddenly is that relevant hardware can be instructed to complete by program, the program can store in a kind of computer-readable storage medium
In matter, which when being executed, includes the steps that one or a combination set of embodiment of the method.
In addition, each functional unit in various embodiments of the present invention can integrate in a processing module, it can also
To be that each unit physically exists alone, can also be integrated in two or more units in a module.It is above-mentioned integrated
Module both can take the form of hardware realization, can also be realized in the form of software function module.The integrated module
If in the form of software function module realize and when sold or used as an independent product, also can store one calculating
In machine read/write memory medium.
Storage medium mentioned above can be read-only memory, disk or CD etc..
Although the embodiments of the present invention has been shown and described above, it is to be understood that above-described embodiment is example
Property, it is not considered as limiting the invention, those skilled in the art within the scope of the invention can be to above-mentioned
Embodiment is changed, modifies, replacement and variant.
Claims (10)
1. a kind of image processing method that panorama is parked characterized by comprising
The multiple images in all angulars field of view of vehicle periphery are obtained using multiple cameras, adjacent two in the multiple camera
A camera acquired image has certain overlapping region;
Described multiple images are handled to obtain multiple birds-eye views respectively;
The characteristic point of the overlapping region of the multiple birds-eye view is extracted respectively and characteristic matching is carried out to the characteristic point, and by institute
It states multiple birds-eye views and projects to the same view visual angle;
The optimal stitching line of the overlapping region in the multiple birds-eye view is established, and will be described according to the optimal stitching line
Multiple birds-eye views are spliced to obtain the first panorama sketch;With
Graph cut is carried out to first panorama sketch to eliminate splicing trace to generate the second panorama sketch.
2. the image processing method that panorama as described in claim 1 is parked, which is characterized in that the camera is flake camera shooting
Head, the processing described multiple images include: to obtain multiple birds-eye views respectively
The multiple fish-eye camera is demarcated, the correction parameter of each fish-eye camera is obtained and utilizes the school
Positive parameter processing described multiple images to obtain multiple correction charts respectively;
According to homography matrix, the multiple correction chart is handled using bilinear interpolation the multiple to obtain respectively
Birds-eye view.
3. the image processing method that panorama as described in claim 1 is parked, which is characterized in that described to establish the multiple get a bird's eye view
The optimal stitching line of the overlapping region in figure includes:
Using each pixel of the frame of the overlapping region as the starting point of suture;
Multiple sutures are determined from the starting point of the suture, and calculate the total intensity value of the pixel of each suture;
With
Select the smallest suture of total intensity value as the optimal stitching line from the multiple suture.
4. the image processing method that panorama as claimed in claim 3 is parked, which is characterized in that described from the suture
Point determines that multiple sutures include: to search adjacent with the starting point from the starting point and be located in next pixel column of the starting point
Current point of the smallest pixel of intensity value as the suture searches adjacent with the current point and is located at the current point
Next pixel column in the smallest pixel of intensity value as the suture next point until exceed the overlapping region.
5. the image processing method that panorama as claimed in claim 3 is parked, which is characterized in that the intensity of each pixel
It is worth calculation formula are as follows:
E (x, y)=Ecolor(x,y)2+Egeometry(x,y);
Wherein, E (x, y) indicates same seat in the overlapping region of two birds-eye views obtained by the two neighboring camera
The intensity value of pixel at mark, Ecolor(x, y) indicates two birds-eye views obtained by the two neighboring camera
The difference of the color value of pixel in overlapping region at same coordinate, Egeometry(x, y) is indicated by the two neighboring camera
The structure difference of pixel in the overlapping region of two birds-eye views obtained at same coordinate.
6. the image processing method that panorama as described in claim 1 is parked, which is characterized in that definition is located at described best
The overlapping region part of suture two sides is respectively the first sub- overlapping region and the second sub- overlapping region, the first son weight
Described image where folded region and the described image where the described second sub- overlapping region belong to two neighboring camera and obtain
Two birds-eye views, it is described to be spliced the multiple birds-eye view according to the optimal stitching line to obtain the first panorama sketch packet
It includes: filling the described first sub- overlapping region and the second sub- overlapping region respectively in the two sides of the optimal stitching line to splice
Obtain first panorama sketch.
7. a kind of panorama parking apparatus characterized by comprising
Multiple cameras, the multiple camera is used to obtain the multiple images in all angulars field of view of vehicle periphery, described more
Two neighboring camera acquired image has certain overlapping region in a camera;
Image processing unit, described image processing unit is for handling described multiple images to obtain multiple birds-eye views respectively;
Image mosaic unit, described image concatenation unit are used to extract the characteristic point of the overlapping region of the multiple birds-eye view respectively
And characteristic matching is carried out to the characteristic point, and the multiple birds-eye view is projected into the same view visual angle, and described in foundation
The optimal stitching line of the overlapping region in multiple birds-eye views, and according to the optimal stitching line by the multiple birds-eye view into
Row splicing obtains the first panorama sketch;With
Seam integrated unit, the seam integrated unit are used to carry out graph cut to first panorama sketch to eliminate splicing trace
Mark is to generate the second panorama sketch.
8. panorama parking apparatus as claimed in claim 7, which is characterized in that the camera is fish-eye camera, the figure
As processing unit includes:
Distortion correction unit, the distortion correction unit obtain each institute for demarcating to the multiple fish-eye camera
State the correction parameter of fish-eye camera and using correction parameter processing described multiple images to obtain multiple correction charts respectively;
With
Birds-eye view processing unit, the birds-eye view processing unit are used for according to homography matrix, using bilinear interpolation to institute
Multiple correction charts are stated to be handled to obtain the multiple birds-eye view respectively.
9. panorama parking apparatus as claimed in claim 7, which is characterized in that described image concatenation unit is used for:
Using each pixel of the frame of the overlapping region as the starting point of suture;
Multiple sutures are determined from the starting point of the suture, and calculate the total intensity value of the pixel of each suture;
With
Select the smallest suture of total intensity value as the optimal stitching line from the multiple suture.
10. panorama parking apparatus as claimed in claim 9, which is characterized in that the intensity value of each pixel calculates public
Formula are as follows:
E (x, y)=Ecolor(x,y)2+Egeometry(x,y);
Wherein, E (x, y) indicates same seat in the overlapping region of two birds-eye views obtained by the two neighboring camera
The intensity value of pixel at mark, Ecolor(x, y) indicates two birds-eye views obtained by the two neighboring camera
The difference of the color value of pixel in overlapping region at same coordinate, Egeometry(x, y) is indicated by the two neighboring camera
The structure difference of pixel in the overlapping region of two birds-eye views obtained at same coordinate.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711387606.5A CN109948398B (en) | 2017-12-20 | 2017-12-20 | Image processing method for panoramic parking and panoramic parking device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711387606.5A CN109948398B (en) | 2017-12-20 | 2017-12-20 | Image processing method for panoramic parking and panoramic parking device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109948398A true CN109948398A (en) | 2019-06-28 |
CN109948398B CN109948398B (en) | 2024-02-13 |
Family
ID=67004349
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711387606.5A Active CN109948398B (en) | 2017-12-20 | 2017-12-20 | Image processing method for panoramic parking and panoramic parking device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109948398B (en) |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110390640A (en) * | 2019-07-29 | 2019-10-29 | 齐鲁工业大学 | Graph cut image split-joint method, system, equipment and medium based on template |
CN110415550A (en) * | 2019-07-31 | 2019-11-05 | 北京智行者科技有限公司 | The automatic parking method of view-based access control model |
CN110443771A (en) * | 2019-08-16 | 2019-11-12 | 同济大学 | It is vehicle-mounted to look around panoramic view brightness and colour consistency method of adjustment in camera system |
CN110555797A (en) * | 2019-08-01 | 2019-12-10 | 天津大学 | panoramic aerial view image illumination homogenization processing method based on least square method |
CN110620874A (en) * | 2019-09-24 | 2019-12-27 | 北京智行者科技有限公司 | Image processing method for parallel driving |
CN111369423A (en) * | 2020-02-05 | 2020-07-03 | 天津大学 | YUV domain panoramic aerial view illumination homogenization processing method based on least square method |
CN111915483A (en) * | 2020-06-24 | 2020-11-10 | 北京迈格威科技有限公司 | Image splicing method and device, computer equipment and storage medium |
CN113379605A (en) * | 2021-07-13 | 2021-09-10 | 广州三目电子有限公司 | Vehicle-mounted 360-degree panoramic image system and computer storage medium |
CN113961078A (en) * | 2021-11-04 | 2022-01-21 | 中国科学院计算机网络信息中心 | Panoramic roaming method, device and equipment and readable storage medium |
WO2023231435A1 (en) * | 2022-06-01 | 2023-12-07 | 北京地平线信息技术有限公司 | Visual perception method and apparatus, and storage medium and electronic device |
CN117237237A (en) * | 2023-11-13 | 2023-12-15 | 深圳元戎启行科技有限公司 | Luminosity balancing method and device for vehicle-mounted 360-degree panoramic image |
CN117455792A (en) * | 2023-12-25 | 2024-01-26 | 武汉车凌智联科技有限公司 | Method for synthesizing and processing 360-degree panoramic image built-in vehicle |
Citations (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102881016A (en) * | 2012-09-19 | 2013-01-16 | 中科院微电子研究所昆山分所 | Vehicle 360-degree surrounding reconstruction method based on internet of vehicles |
CN103533266A (en) * | 2013-10-01 | 2014-01-22 | 中国人民解放军国防科学技术大学 | 360-degree stitched-type panoramic camera with wide view field in vertical direction |
CN103863192A (en) * | 2014-04-03 | 2014-06-18 | 深圳市德赛微电子技术有限公司 | Method and system for vehicle-mounted panoramic imaging assistance |
CN203759754U (en) * | 2014-04-03 | 2014-08-06 | 深圳市德赛微电子技术有限公司 | Vehicular panoramic imaging system calibrator |
CN105245841A (en) * | 2015-10-08 | 2016-01-13 | 北京工业大学 | CUDA (Compute Unified Device Architecture)-based panoramic video monitoring system |
CN105447850A (en) * | 2015-11-12 | 2016-03-30 | 浙江大学 | Panorama stitching synthesis method based on multi-view images |
CN105513031A (en) * | 2014-10-14 | 2016-04-20 | 清华大学 | Graph matching-based data driving image extrapolation method and device |
CN105654502A (en) * | 2016-03-30 | 2016-06-08 | 广州市盛光微电子有限公司 | Panorama camera calibration device and method based on multiple lenses and multiple sensors |
CN105894451A (en) * | 2016-03-30 | 2016-08-24 | 沈阳泰科易科技有限公司 | Method and device for splicing panoramic image |
CN106204437A (en) * | 2016-06-28 | 2016-12-07 | 深圳市凌云视迅科技有限责任公司 | A kind of image interfusion method |
CN106530267A (en) * | 2016-11-30 | 2017-03-22 | 长沙全度影像科技有限公司 | Fusion method for avoiding panoramic picture misalignment |
CN106780344A (en) * | 2017-01-06 | 2017-05-31 | 苏州大学 | 360 ° of the automobile for being shifted based on deviation and being split looks around image split-joint method |
CN106952311A (en) * | 2017-03-02 | 2017-07-14 | 山东省科学院自动化研究所 | Auxiliary parking system and method based on panoramic mosaic data mapping tables |
CN206348849U (en) * | 2017-01-09 | 2017-07-21 | 上海商泰汽车信息系统有限公司 | Scaling board |
CN107004277A (en) * | 2014-12-04 | 2017-08-01 | 康诺特电子有限公司 | The on-line proving of motor vehicles camera system |
-
2017
- 2017-12-20 CN CN201711387606.5A patent/CN109948398B/en active Active
Patent Citations (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102881016A (en) * | 2012-09-19 | 2013-01-16 | 中科院微电子研究所昆山分所 | Vehicle 360-degree surrounding reconstruction method based on internet of vehicles |
CN103533266A (en) * | 2013-10-01 | 2014-01-22 | 中国人民解放军国防科学技术大学 | 360-degree stitched-type panoramic camera with wide view field in vertical direction |
CN103863192A (en) * | 2014-04-03 | 2014-06-18 | 深圳市德赛微电子技术有限公司 | Method and system for vehicle-mounted panoramic imaging assistance |
CN203759754U (en) * | 2014-04-03 | 2014-08-06 | 深圳市德赛微电子技术有限公司 | Vehicular panoramic imaging system calibrator |
CN105513031A (en) * | 2014-10-14 | 2016-04-20 | 清华大学 | Graph matching-based data driving image extrapolation method and device |
CN107004277A (en) * | 2014-12-04 | 2017-08-01 | 康诺特电子有限公司 | The on-line proving of motor vehicles camera system |
CN105245841A (en) * | 2015-10-08 | 2016-01-13 | 北京工业大学 | CUDA (Compute Unified Device Architecture)-based panoramic video monitoring system |
CN105447850A (en) * | 2015-11-12 | 2016-03-30 | 浙江大学 | Panorama stitching synthesis method based on multi-view images |
CN105654502A (en) * | 2016-03-30 | 2016-06-08 | 广州市盛光微电子有限公司 | Panorama camera calibration device and method based on multiple lenses and multiple sensors |
CN105894451A (en) * | 2016-03-30 | 2016-08-24 | 沈阳泰科易科技有限公司 | Method and device for splicing panoramic image |
CN106204437A (en) * | 2016-06-28 | 2016-12-07 | 深圳市凌云视迅科技有限责任公司 | A kind of image interfusion method |
CN106530267A (en) * | 2016-11-30 | 2017-03-22 | 长沙全度影像科技有限公司 | Fusion method for avoiding panoramic picture misalignment |
CN106780344A (en) * | 2017-01-06 | 2017-05-31 | 苏州大学 | 360 ° of the automobile for being shifted based on deviation and being split looks around image split-joint method |
CN206348849U (en) * | 2017-01-09 | 2017-07-21 | 上海商泰汽车信息系统有限公司 | Scaling board |
CN106952311A (en) * | 2017-03-02 | 2017-07-14 | 山东省科学院自动化研究所 | Auxiliary parking system and method based on panoramic mosaic data mapping tables |
Non-Patent Citations (1)
Title |
---|
刘凤英;陈天恩;王冬;: "多影像空三平差辅助下的车载全景相机严密标定方法", 测绘通报, no. 07 * |
Cited By (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110390640A (en) * | 2019-07-29 | 2019-10-29 | 齐鲁工业大学 | Graph cut image split-joint method, system, equipment and medium based on template |
CN110415550A (en) * | 2019-07-31 | 2019-11-05 | 北京智行者科技有限公司 | The automatic parking method of view-based access control model |
CN110555797B (en) * | 2019-08-01 | 2023-04-25 | 天津大学 | Panoramic aerial view image illumination homogenization processing method based on least square method |
CN110555797A (en) * | 2019-08-01 | 2019-12-10 | 天津大学 | panoramic aerial view image illumination homogenization processing method based on least square method |
CN110443771A (en) * | 2019-08-16 | 2019-11-12 | 同济大学 | It is vehicle-mounted to look around panoramic view brightness and colour consistency method of adjustment in camera system |
CN110443771B (en) * | 2019-08-16 | 2023-07-21 | 同济大学 | Method for adjusting consistency of brightness and color of annular view in vehicle-mounted annular view camera system |
CN110620874A (en) * | 2019-09-24 | 2019-12-27 | 北京智行者科技有限公司 | Image processing method for parallel driving |
CN110620874B (en) * | 2019-09-24 | 2021-09-10 | 北京智行者科技有限公司 | Image processing method for parallel driving |
CN111369423A (en) * | 2020-02-05 | 2020-07-03 | 天津大学 | YUV domain panoramic aerial view illumination homogenization processing method based on least square method |
WO2021258579A1 (en) * | 2020-06-24 | 2021-12-30 | 北京迈格威科技有限公司 | Image splicing method and apparatus, computer device, and storage medium |
CN111915483A (en) * | 2020-06-24 | 2020-11-10 | 北京迈格威科技有限公司 | Image splicing method and device, computer equipment and storage medium |
CN111915483B (en) * | 2020-06-24 | 2024-03-19 | 北京迈格威科技有限公司 | Image stitching method, device, computer equipment and storage medium |
CN113379605A (en) * | 2021-07-13 | 2021-09-10 | 广州三目电子有限公司 | Vehicle-mounted 360-degree panoramic image system and computer storage medium |
CN113379605B (en) * | 2021-07-13 | 2021-12-31 | 广州三目电子有限公司 | Vehicle-mounted 360-degree panoramic image system and computer storage medium |
CN113961078A (en) * | 2021-11-04 | 2022-01-21 | 中国科学院计算机网络信息中心 | Panoramic roaming method, device and equipment and readable storage medium |
WO2023231435A1 (en) * | 2022-06-01 | 2023-12-07 | 北京地平线信息技术有限公司 | Visual perception method and apparatus, and storage medium and electronic device |
CN117237237A (en) * | 2023-11-13 | 2023-12-15 | 深圳元戎启行科技有限公司 | Luminosity balancing method and device for vehicle-mounted 360-degree panoramic image |
CN117455792A (en) * | 2023-12-25 | 2024-01-26 | 武汉车凌智联科技有限公司 | Method for synthesizing and processing 360-degree panoramic image built-in vehicle |
CN117455792B (en) * | 2023-12-25 | 2024-03-22 | 武汉车凌智联科技有限公司 | Method for synthesizing and processing 360-degree panoramic image built-in vehicle |
Also Published As
Publication number | Publication date |
---|---|
CN109948398B (en) | 2024-02-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109948398A (en) | The image processing method and panorama parking apparatus that panorama is parked | |
CN104851076B (en) | Panoramic looking-around parking assisting system and camera installation method for commercial car | |
CN112224132B (en) | Vehicle panoramic all-around obstacle early warning method | |
CN103617606B (en) | For assisting the vehicle multi-angle panorama generation method of driving | |
JP2022095776A (en) | Rear-stitched view panorama for rear-view visualization | |
JP5739584B2 (en) | 3D image synthesizing apparatus and method for visualizing vehicle periphery | |
US8514282B2 (en) | Vehicle periphery display device and method for vehicle periphery image | |
US8928753B2 (en) | Method and apparatus for generating a surrounding image | |
CN108263283A (en) | More marshalling varied angle vehicle panoramic viewing system calibration and joining method | |
JP5077606B2 (en) | Method for calibrating a device comprising at least one omnidirectional camera and an optical display unit | |
Yu et al. | 360 surround view system with parking guidance | |
JP2009105656A (en) | On-vehicle imaging apparatus | |
JP2008077628A (en) | Image processor and vehicle surrounding visual field support device and method | |
KR20190047027A (en) | How to provide a rearview mirror view of the vehicle's surroundings in the vehicle | |
CN104057882A (en) | System For Viewing A Curb In A Front Region On The Basis Of Two Cameras | |
CN112070886B (en) | Image monitoring method and related equipment for mining dump truck | |
JP4830380B2 (en) | Vehicle periphery monitoring device and vehicle periphery monitoring method | |
CN109087251A (en) | A kind of vehicle-mounted panoramic image display method and system | |
US11528453B2 (en) | Sensor fusion based perceptually enhanced surround view | |
JP2009105660A (en) | On-vehicle imaging apparatus | |
CN112215747A (en) | Method and device for generating vehicle-mounted panoramic picture without vehicle bottom blind area and storage medium | |
JP5405809B2 (en) | Vehicle periphery monitoring device and camera position / posture determination method | |
TWM524490U (en) | Driving panoramic auxiliary device | |
JP6293089B2 (en) | Rear monitor | |
CN112308986B (en) | Vehicle-mounted image stitching method, system and device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |